The key hypothesis of the proposed component is that the network(s) associated with consumption/withdrawal are genotype independent, and thus would emerge regardless of the specific cross examined. To test this hypothesis we will use reciprocal short-term selective breeding (rSTSB) in three different mouse populations: a C57BL/6J (B6) x DBA/2J (D2) F2 intercross, HS4 animals (formed by crossing the B6, D2, BALB/cJ (C) and LP/J (LP) inbred mouse strains) followed by 20+ generations of intercrossing, and the HS-CC, an outbred version of the collaborative cross mouse (Churchill et al. 2004). rSTSB, as used here, involves selection of one line for high ethanol withdrawal &low ethanol consumption and vice versa for the second oppositely-selected line. The specific aims may be summarized: 1. To use rSTSB to select lines of mice for high consumption and low withdrawal and vice versa. One substantial advantage of rSTSB is the reduction in the number of lines needed for the network analyses by 50% (when compared to separate selections for consumption and withdrawal). 2. To examine the paired reciprocal lines for differential gene expression and weighted gene co-expression network analysis. Gene expression will be assessed for each pair of rSTSB lines using the lllumina Mouse WG6 2.0 array. Gene expression will be initially measured in the central nucleus of the amygdala (CeA) given the evidence that the CeA has a key role in both phenotypes of interest (Dhaher et al. 2008;Chen et al. 2008;2009). As the network analyses are repeated for each selection, it is expected that common gene networks and/or key regulatory factors will emerge. However, the selection experiments also allow one to test an alternative hypothesis for selection;thus, we propose.3. To determine the extent that non-synonymous coding SNPs segregate in the rSTSB lines. In depth Next-Generation transcriptome sequencing will be used to assess the extent of Cn SNP segregation which in turn will generate a gene list that will be analyzed using the same general approach as for the differentially expressed transcripts. As with the gene expression data, the value of the Cn SNP data is based on the detection of similar networks across the three different genotypes.